DATASET_Methodological aspects of water eDNA filtering for vertebrate biodiversity assessments in Mediterranean ecosystems.

Global biodiversity loss driven by ecosystem degradation underscores the need for standardized, reliable methodologies to monitor community change and inform conservation action. Environmental DNA (eDNA) metabarcoding has emerged as a versatile and scalable tool for biodiversity assessment; however, limited methodological standardization constrains comparability and reproducibility across studies. In this study, we evaluated how methodological choices, from field sampling to laboratory processing and bioinformatic analysis, affect the characterization of vertebrate richness and community composition using a capsule-based water filtration system. Specifically, we sampled six lentic water bodies within Mediterranean ecosystems and tested how capsule pore size, biological replication and pooling strategies, molecular marker selection, and library preparation and cleaning procedures influence estimates of vertebrate diversity and community composition. Our results showed that detection of vertebrate taxa in lentic systems seems to be optimized by using 0.45 µm capsule filters, independently analysing PCR replicates, and pooling sequencing data post hoc. The combined use of multiple molecular markers further improved taxonomic coverage. Library preparation and data processing approaches also had an influence. Overall, our findings demonstrate that methodological aspects substantially influence eDNA-based biodiversity monitoring. Systematic evaluation and harmonization of protocols are therefore essential to enhance comparability among studies, maximize taxonomic detection, and strengthen the application of eDNA metabarcoding in long-term vertebrate monitoring programs. Keywords: hydric resources; integrated wildlife monitoring; molecular ecology; terrestrial mammals; waterholes; filtration. VARIABLES Capsule pore size: 22: capsules with 0.22 µm mesh. 45: capsules with 0.45 µm mesh. Capsule replicate: 1 & 2. Test:Standard: DNA extracted from each capsule was processed independently through PCR1 and PCR2, resulting in 24 samples per marker (6 sites × 4 capsules), with three PCR replicates (total = 72 libraries). Pooled (tested with Vert01 marker only): DNA extracts were first processed independently through PCR1. PCR2 was then performed in a larger reaction volume (40 µL), followed by three alternative clean-up strategies. Test 1: individual clean-up of each PCR2 product (12 samples × 3 PCR replicates = 36 libraries); Test 2a: pooled clean-up of 3 PCR replicates using 12.5 µL per sample (24 libraries); Test 2b: pooled clean-up of 3 PCR replicates using 5 µL per sample (24 libraries). Standard2Caps: DNA extracts from capsules of the same pore size within each site were pooled prior to PCR1. The pooled samples were then processed through PCR1 and PCR2, resulting in 12 samples per marker (6 sites × 2 pore sizes), with three PCR replicates (total = 36 libraries). No capsule replicates: number 3 assigned. Markers: Vert: designed to detect vertebrates. Mamm: designed to detect mammals. Taxonomic assignment threshold (TAT; %): 97 & 99.